ANTICANCER RESEARCH 34: 4539-4550 (2014)

Multicolor Analysis of Cell Surface Marker of Human Leukemia Cell Lines Using Flow Cytometry

TOMOKO INOUE1, ANTHONY SWAIN1, YOICHI NAKANISHI2 and DAISUKE SUGIYAMA1

1Department of Research and Development of Next-Generation Medicine, Faculty of Medical Sciences, Kyushu University, Fukuoka, Japan; 2Center for Clinical and Translational Research, Kyushu University Hospital, Fukuoka, Japan

Abstract. Background: Leukemia cell lines are utilized as and transcription factor alterations, which lead to anemia, tools for molecular analysis. Their implementation in therapy neutropenia and thrombocytopenia. In almost all cases, will require standards for quality control, including leukemia is classified as one of four types; acute appropriate selection criteria for functional analysis and myelogenous leukemia (AML), acute lymphocytic leukemia efficacy determination. Materials and Methods: (ALL), chronic myelogenous leukemia (CML), and chronic Characteristics of six human leukemia cell lines –Kasumi-1, lymphocytic leukemia (CLL). In acute leukemia, leukocytes NB-4, MOLM-13, MV-4-11, K562, and Jurkat cells– were proliferate rapidly in the bone marrow at an immature state, investigated using multiple color analysis of surface antigen and therefore cannot function normally. On the other hand, in expression and comparative analysis of gene expression. chronic leukemia, blastic cells form more slowly than in Results: Differentiation states of Kasumi-1 and MOLM-13 acute leukemia and consequently cells with abnormal cells are colony-forming units-granulocyte/macrophage function are transported to the hematopoietic tissues. In equivalent cells to myeloblasts with comparatively high AML and CML, myeloid lineage cells such as granulocytes Growth factor independent-1(GFI1) and Transcription factor and monocytes, become malignant and lose the capacity to PU.1 (PU.1) expression, respectively. NB4 and MV-4-11 differentiate into mature functional cells and interfere with express high levels of CCAAT/enhancer-binding protein- the production of normal blood cells. In ALL and CLL, alpha (CEBPα) and differentiate from myeloblasts to pro- lymphoid lineage cells such as T- and B-lymphocytes do monocytes and myeloblasts, respectively. K562 cells are likewise. There is considerable research taking place utilizing colony-forming units-erythroid equivalent cells to leukemia cells (both primary cells and immortalized cell erythroblasts, with the highest expression of GATA-binding lines) to elucidate their pathogenetic mechanisms at cellular factor 2 (GATA2), GATA1 and Friend of gata-1 (FOG1). and molecular levels, and to develop novel therapies (for Jurkat cells are pro-T to mature T-cells with the highest example, gene therapy and molecular-target therapy) in Neurogenic locus notch-1 homolog protein 1 (NOTCH1) addition to existing therapeutic methods, such as expression. Conclusion: Our study gives a useful guideline chemotherapy, hematopoietic stem cell transplantation, of standards for appropriate usage of leukemia cell lines for transfusion therapy and induction therapy. examining novel targets in vitro. Primary cells from patients with leukemia are obtained from peripheral blood (PB) and bone marrow (BM), which provide Leukemias are a group of disorders characterized by much information for diagnosis through morphological abnormal growth and differentiation of hematopoietic cells observation, cell surface antigen-dependent phenotypical due to several factors, including chromosomal abnormalities observations and chromosomal tests. However, as primary cells are limited in number and accessibility (especially for BM), it is not easy to extract them for functional analysis in order to examine novel targets to treat patients. To overcome this Correspondence to: Dr. Tomoko Inoue and Dr. Daisuke Sugiyama, problem, immortalized leukemia cell lines that can proliferate Department of Research and Development of Next-Generation indefinitely were established and are being utilized as useful Medicine, Faculty of Medical Sciences, Kyushu University, 3-1-1, tools to analyze molecular mechanisms (transcription factors, Maidashi, Higashi-Ku, Fukuoka, 812-8582, Japan. Tel: +81 926426210, Fax: +81 926426146, e-mail: [email protected] signaling pathways, translocations) (1, 2). They are also used (Tomoko Inoue), [email protected] (Daisuke Sugiyama) for analysis of both gain-of-function, with plasmid transfection and use of recombinant protein, and loss-of-function, with Key Words: Leukemia cell line, cell surface marker, flow cytometry. siRNA/shRNA transfection and inhibitors (3, 4).

0250-7005/2014 $2.00+.40 4539 ANTICANCER RESEARCH 34: 4539-4550 (2014)

When the cell lines were established, expressions of cell To analyze myeloid cell lineage markers, cells were stained with a surface markers were studied in terms of single antigens but FITC-conjugated anti-human CD15 (Biolegend), PE-conjugated anti- not multiple antigens together. The transcription factors were human CD116 (Biolegend), PE-Cy7-conjugated anti-human CD14 (Biolegend), APC-conjugated anti-human CD11c (Biolegend), APC- examined individually and not compared to other cell lines. Cy7-conjugated anti-human CD13 (Biolegend), Pacific Blue It is necessary to characterize leukemia cell lines in terms of conjugated anti-mouse/human CD11b (Biolegend), Biotin-conjugated protein and gene expression levels, so as to evaluate the state anti-human CD16 (Biolegend) and V500-conjugated streptavidin (BD of differentiation and maturation more precisely, which will Bioscience, San Jose, CA, USA). be very useful in controlling the quality of cell lines, as well To analyze lymphoid cell lineage markers, cells were stained with as in choosing cell lines to evaluate the effects of molecules a FITC-conjugated anti-human CD19 (Biolegend), PE-conjugated and therapies. anti-human CD2 (Biolegend), APC-conjugated anti-human CD3 (Biolegend) and APC-Cy7-conjugated anti-human CD4 (Biolegend). Multi-parametric flow cytometry is an adequate method MOLM-13 and MV-4-11 cells: To analyze HSPC markers, cells for gaining information on the cell differentiation state, were stained with a FITC-conjugated anti-human CD38, PE- monitoring residual disease and evaluating response to conjugated anti-human CD33, PE-Cy7-conjugated anti-human c-KIT, therapy (5, 6). Such application of flow cytometry has APC-conjugated anti-human CD34 and Pacific Blue conjugated anti- become more widespread, and will require the development human HLA-DR. of standardized approaches with defined specificity and To analyze myeloid cell lineage markers, cells were stained with sensitivity, along with suitable quality control schemes. a FITC-conjugated anti-human CD15, PE-conjugated anti- mouse/human CD11b (Biolegend), PE-Cy7-conjugated anti-human Herein we examined the multiple color analysis of cell CD14, APC-conjugated anti-human CD36 (Biolegend), APC-Cy7- surface markers and comparative analysis of transcription conjugated anti-human CD13, Pacific Blue conjugated anti- human gene expression in using six human leukemia cell lines: four HLA-DR, Biotin-conjugated anti-human CD64 (Biolegend) and myeloblastic and monoblastic leukemia cell lines: Kasumi-1 V500-conjugated streptavidin. (7), NB-4 (8), MOLM-13 (9) and MV-4-11 (10); an To analyze lymphoid cell lineage markers, cells were stained erythroleukemia cell line, K562 (11); and the T-cell leukemic with a FITC-conjugated anti-human CD10 (Biolegend), PE- cell line Jurkat (12), which have been previously used as in conjugated anti-human CD116, PE-Cy7-conjugated anti-human CD14, APC-conjugated anti-human CD11c Ab, APC-Cy7- vitro leukemia models. conjugated anti-human CD4 (Biolegend), Pacific Blue conjugated anti-human HLA-DR, Biotin-conjugated anti-human CD45 Materials and Methods (Biolegend) and V500-conjugated streptavidin. K562 cells: To analyze HSPC markers, cells were stained with a Cell culture. The human leukemia cell lines, Kasumi-1, NB4, MOLM- FITC-conjugated anti-human CD38, PE-conjugated anti-human 13, K562 and Jurkat were cultured in RPMI-1640 medium (Wako Pure CD33, PE-Cy7-conjugated anti-human c-KIT, APC-conjugated anti- Chemical Industries, Osaka, Japan) supplemented with 10% fetal human CD34, APC-Cy7-conjugated anti-human CD13, and Pacific bovine serum and MV-4-11 cells were cultured in RPMI-1640 medium Blue conjugated anti-human HLA-DR. supplemented with 10% fetal bovine serum and 10 ng/ml human To analyze erythroid cell lineage markers, cells were stained with recombinant Granulocyte/macrophage colony-stimulating factor (GM- a FITC-conjugated anti-human Glycophorin A (GPA, CD235a) CSF) (PeproTech, Rochy Hill, NJ, USA) at 37˚C in 5% CO2. (Biolegend), PE-conjugated anti- human CD71 (Biolegend), PE- Cy7-conjugated anti-human c-KIT, APC-conjugated anti-human Cell morphology. Slides of cell suspensions were prepared by CytoSpin CD36 (Biolegend), and APC-Cy7-conjugated anti-mouse/human 4 (Thermo Fisher scientific, Waltham, MA, USA) at 450 rpm (rcf, CD44 (Biolegend). 26.06 × g) for 7 minutes. Cells were stained with May-Grunwald and Jurkat cells: To analyze HSPC markers, cells were stained with Giemsa staining reagents (Muto Pure Chemicals, Tokyo, Japan) and FITC-conjugated anti-human CD38, PE-conjugated anti-human characterized morphologically. One slide containing 2×104 cells was CD33, PE-Cy7-conjugated anti-human c-KIT, and APC-conjugated prepared for each cell line. Stained cells were observed using an anti-human CD34. Olympus CKX41 microscope (Olympus, Tokyo, Japan). To analyze lymphoid cell lineage markers, cells were stained with FITC-conjugated anti-human CD7 (Biolegend), PE-conjugated Flow cytometry. Kasumi-1 cells and NB4 cells: To analyze anti-human CD2 (Biolegend), PE-Cy7-conjugated anti-human CD8 hematopoietic stem/progenitor cell (HSPC) markers, cells were stained (Biolegend), APC-conjugated anti-human CD3 (Biolegend), APC- with a Fluorescein isothiocyanate (FITC)-conjugated anti-human Cy7-conjugated anti-mouse/human CD4 (Biolegend), Pacific Blue Cluster of differentiation (CD)38 (Biolegend, San Diego, CA, USA), conjugated anti-human CD5 (Biolegend), Biotin-conjugated anti- Phycoerythrin (PE)-Cy7-conjugated anti-human Tyrosine-protein human CD38 and V500-conjugated streptavidin. kinase (c-KIT, CD117) (Biolegend), Allophycocyanin (APC)- The cells were analyzed using a FACS Aria cell sorter (BDIS, conjugated anti-human CD34 (eBioscience, San Diego, CA, USA), San Jose, CA, USA), and the data files were analyzed using FlowJo APC-Cy7-conjugated anti-human CD13 (Biolegend) and Pacific Blue- software (Tree Star, Inc., Sac Carlos, CA, USA). Data are presented conjugated anti-human leukocyte antigen (HLA)-DR (Biolegend). In as means plus standard deviation (SD). addition, PE-conjugated anti-human CD135 (Biolegend) for Kasumi cells and PE-conjugated anti-human CD33 (Biolegend) for NB4 cells Quantitative real-time reverse transcription–polymerase chain were stained to analyze the cells. reaction (qRT-PCR). Total RNA was extracted from each cell line

4540 Inoue et al: Multicolor Analysis of Cell Surface Markers of Human Leukemia Cell Lines

Table I. Human leukemia cell lines

Cell lines Cell source Type Reference

Kasumi-1 Peripheral blood Acute myeloblastic leukemia (M2) 7 NB4 Bone marrow Acute promyelocytic leukemia (M3) 8 MOLM-13 Peripheral blood Acute monocytic leukemia (M5a) 9 MV-4-11 Bone marrow/peripheral blood Childhood acute myeloblastic leukemia (M5) 10 K562 Bone marrow Chronic myelogenous leukemia/erythroleukemia 11 Jurkat Peripheral blood Acute T-cell leukemia 12

using an RNAqueous Micro Kit (Life Technologies, Carlsbad, CA, all cell lines; myeloid-specific markers such as CD45, CD13, USA). Total RNA was subjected to reverse transcription using a CD14, CD15 for Kasumi-1, NB4, MOLM-13 and MV-4-11; High-Capacity RNA-to-cDNA Kit (Life Technologies) according to erythroid-specific markers such as GPA, CD36, CD71 for established protocols. The mRNA levels of different genes were analyzed by qRT-PCR using SYBR Green and gene-specific primers K562 cells; and T-lymphocyte specific markers such as CD3, with the StepOnePlus real-time PCR system (Applied Biosystems, CD4, CD8 for Jurkat cells. Foster City, CA, USA). The mRNA level of each target gene was The percentages of positive cells for each marker are shown normalized to β-actin (ACTB) as an internal control. in Figure 2. The myelo/monoblastic leukemia cell lines, Kasumi-1, NB4, MOLM-13, and MV-4-11 were mostly Results positive for CD33 (HSPC marker) and CD45 (pan-leucocyte marker), in 99.8 to 100% and 98.9 to 99.6% of cells, Morphological observation of human leukemia cell lines. respectively. Among them, Kasumi-1 cells were likely most Leukemia cell lines used in this study are shown in Table I. immature, defined by positivity for c-KIT (99.8%) and CD34 To examine cell morphology, alive Propidium iodide (PI)- (95.5%), and also expressed CD135 (2.2%). On the other negative cells were sorted-out and stained with May-Grünwald hand, few NB4, MOLM and MV-4-11 cells were positive for Giemsa solution (Figure 1). Typical cell characteristics were CD34 (0.17 to 0.34%) and showed a variety of expression of observed under microscopy. Each cell line showed variations CD38 (56.5%, 94.3%, 0.68%, respectively) and HLA-DR in size, nuclear segmentation, nuclear:cytoplasmic (N/C) ratio, (19.6%, 23.4%, 85.0%, respectively). Regarding and content of cytoplasm, such as granules and vacuoles. differentiation marker CD13 (pan-myeloid specific marker, Kasumi cells (20.95±2.58 μm) showed nuclear segmentation widely expressed from Colony formings-granulocyte/ (22.2%) in nuclei and vacuoles in the cytoplasm (85.7%). macrophage (CFU-GM) equivalent cells to monocytes), NB4 Azurophilic granules were contained inside the basophilic cells were the ones most positive (99.8%), followed by cytoplasm (Figure 1, upper left). No nuclear segmentation was Kasumi-1 cells (44.0%), MV-4-11 (5.38%) and MOLM-13 observed in NB4 cells (24.49±2.91 μm) whose cytoplasm (1.32%) cells. NB4 cells showed the highest positivity for contained vacuoles (42.9%) with high N/C ratio. Azurophilic CD14 (30.4%), CD11b (80.4%) and CD11c (69.2%), and MV- granules were contained in the basophilic cytoplasm (Figure 4-11 did for CD15 (88.6%) and CD64 (99.6%) among the four 1, lower left). MOLM-13 cells (25.10±3.18 μm) did not have cell lines (Figure 2). Erythroleukemia K562 cells had both nuclear segmentation or vacuoles in the cytoplasm (42.8%). immature state defined by c-KIT (12.9%) and CD33 (61.7%) Azurophilic granules were contained in the basophilic expression and mature state defined by CD36 (75.3%), CD71 cytoplasm (Figure 1, upper center). Nuclear condensation was (99.8%) and GPA (19.7%). T-cell leukemia Jurkat cells were observed in MV-4-11 cells (23.57±1.79 μm), but no nuclear the most mature among the six cell lines tested, defined by c- segmentation or vacuoles in their cytoplasm were apparent KIT (1.22%). They were positive for markers of maturity: (Figure 1, lower center). K562 cells (25.12±2.23 μm) also CD7 (pan-T-cell-specific marker, widely expressed from pro- exhibited azurophilic granules in the basophilic cytoplasm T-cell to mature T-cell) in 60.9%, CD2 and CD5 (expressed (Figure 1, upper right). Compared to other cell lines, Jurkat from early thymocyte to mature thymocyte) as 59.0% and cells were smaller in size (17.08±2.38 μm) and had a lower 57.0%, respectively, and CD3 and CD4 (mature T-cell marker) N/C ratio (Figure 1, lower right). in 45.2% and 28.6%, respectively. To further characterize the differentiated state of each cell Surface marker expression of human leukemia cell lines. To line, we analyzed the surface marker expression under further investigate surface marker expression of human combination (Figure 3). All combination analyses are leukemia cell lines, we performed flow cytometric analysis summarized in Table II. Kasumi-1 cells were double-positive using HSPC markers such as c-KIT, CD34, CD33, CD38 for for c-KIT/CD34 at 95.4%, CD33/HLA-DR at 7.4%, and

4541 ANTICANCER RESEARCH 34: 4539-4550 (2014)

Figure 1. Morphology of human leukemia cell lines. Morphology of Kasumi-1 (upper left), NB4 (lower left), MOLM-13 (upper center), MV-4-11 (lower center), K562 (upper right) and Jurkat (lower right) cells. All cell lines were stained with the May-Giemsa staining method. Original magnification, ×40 for all panels; scale bar is 20 μm.

CD13/HLA-DR at 5.2%. Among CD13/CD116 double- CD10/CD13 (Figure 3 D). K562 cells were double positive positive (both are express from CFU-GM to mature for c-KIT/CD33 and c-KIT/CD13 at 8.9% and 2.1%, but monocytes), CD15-positive cells and CD11b-positive cells rarely for CD33/CD38, c-KIT/CD34 or HLA-DR/CD13. were observed at 2.12% and 64.6%, respectively, although Among c-KIT-negative/CD36-positive cells that included we did not see cells highly positive for combinations Colony forming units-erythroid (CFU-E) equivalent cells, CD14/CD15, CD15/CD16, CD11b/CD11c cells (Figure 3A). erythroblasts and mature erythrocytes, CD71/GPA double- NB4 cells were double-positive for c-KIT/CD34, c- positive polychromatic and CD44/GPA double-positive KIT/CD38, CD34/CD38 and c-KIT/HLA-DR cell at 0.06%, orthochromatic erythroblasts at 3.2% and 2.7% (Figure 3E). 14.4%, 2.6% and 2.7%, respectively. Regarding Jurkat cells were rarely double positive for CD33/CD34, c- differentiation markers, positivity for CD15/CD116, KIT/CD34, CD34/CD38 cells and CD38/HLA-DR. Among CD11b/CD11c and CD14/CD11c cell were observed at double positive CD38/CD7 cells that are expected to include 1.96%, 21.2% and 22.4%, respectively (Figure 3B). MOLM- pro-T cell and thymocytes, CD2/CD5, CD3/CD4 and 13 cells were double-positive for c-KIT/CD38 and CD3/CD8 cells existed at 98.7%, 44.8%, and 1.2%. Among CD33/HLA-DR 9.83%, and 25.2%, but rarely for c- CD38-negative/CD7-positive cells that are expected to be KIT/CD34. They were also positive for HLA-DR/CD13 cells mature, functional T-cells, double positive CD2/CD5 and at 2.2%. Compared to other cell lines, they were positive for CD3/CD4 cells were found at 94.6% and 36.0%, but were single-cell markers, such as CD11b, CD14 and CD15, but no negative for CD3/CD8 (Figure 3F). Taken together from double-positivity was observed (Figure 3C). MV-4-11 cells immunophenotype analysis, the differentiation state of were double-positive for CD33/HLA-DR at 85.8%, but rarely Kasumi-1 and MOLM-13 cells appears to be CFU-GM to for c-KIT/CD34 and CD34/CD38. They were also positive myeloblasts, NB4 cells are myeloblasts to pro-monocytes; for HLA-DR/CD13 at 0.9%. Regarding differentiation MV-4-11 cells are myeloblasts; K562 cells are CFU-E to marker, CD15/CD64 were observed at 27.4%, but were erythroblasts; and Jurkat cells are pro-T cells to mature rarely positive for CD14/CD116, CD36/CD63, and thymocytes and functional T-cells.

4542 Inoue et al: Multicolor Analysis of Cell Surface Markers of Human Leukemia Cell Lines

Figure 2. Surface phenotype analysis with flow cytometry. Cells were stained with surface antigens, such as Tyrosine-protein kinase (c-KIT), Cluster of differentiation (CD) 34, CD33, CD38 and CD45, and analyzed with flow cytometry. Data are shown for each individual positive cell marker.

Hematopoietic transcription factor expression in human (7230.01) and GFI1 (5565.11). MOLM-13 had higher leukemia cell lines. Next, we examined expression of expression of PU.1 (11553.68) than GFI1 (3848.21) and low hematopoietic transcription factor genes, such as PU.1, expression of CEBPα (214.20). MV-4-11 predominantly CEBPα, GFI1, NOTCH1, E2A, GATA2, GATA1 and FOG1 in expressed CEBPα (120013.11) and also expressed PU.1 six leukemia cell lines (Figure 4). Myeloblastic/ monoblastic (11364.33) and GFI1 (4338.27). K562 cells predominantly cell lines (Kasumi-1, NB4, MOLM-13, and MV-4-11) express expressed GATA1 (1.2E+09) and also expressed GATA2 myeloid-specific transcription factors PU.1 and GFI1. In (2432.562). Jurkat cells expressed 9.16-fold more NOTCH1 contrast, CEBPα expression differed among cell lines, its than E2A (Figure 4). Overall, we found that the expression of expression in MV-4-11 was the highest and 9.45-fold higher hematopoietic transcription factors varies among leukemia in NB4 cells. Jurkat T-cell leukemia cells more highly cell lines at different levels. expressed NOTCH1 and E2A than did K562 erythroleukemia cell lines, at 8.09-fold and 6.82-fold higher, respectively. When Discussion compared to myeloblastic/monoblastic cell lines, Jurkat had the highest expression of both NOTCH1 and E2A. K562 cells Herein, we carried-out multiple color analysis of cell surface more highly expressed GATA2 and GATA1 than Jurkat cells, markers by flow cytometry, and comparative analysis of at 110.5-fold and 839.4-fold higher, respectively. When expression of transcription genes on six human leukemia cell compared to myeloblastic/monoblastic cell lines, it had the lines that are commonly used for studies on in vitro leukemic highest expression of both GATA2 and GATA1 (Figure 4). models. Our results suggest guidelines of standards for (i) Among cell lines, Kasumi cells expressed GFI1 (7837.28), quality control of cell lines, (ii) choosing appropriate cell PU.1 (6010.33) and low expression of CEBPα (52.69). NB4 lines for functional analysis, and (iii) determination of cells had higher expression of CEBPα (12699.06) than PU.1 efficacy after functional analysis and drug screening.

4543 ANTICANCER RESEARCH 34: 4539-4550 (2014)

Figure 3. Multiple color analysis of surface marker phenotype. Multiple color flow cytometric analyses were performed using Kasumi-1 (A), NB4 (B), MOLM-13 (C), MV-4-11 (D), K562 (E) and Jurkat (F) cells. Three independent flow cytometric analyses were carried-out in all cell lines and typical plot patterns are shown here.

4544 Inoue et al: Multicolor Analysis of Cell Surface Markers of Human Leukemia Cell Lines

Figure 3. Continued

4545 ANTICANCER RESEARCH 34: 4539-4550 (2014)

Figure 3. Continued

4546 Inoue et al: Multicolor Analysis of Cell Surface Markers of Human Leukemia Cell Lines

Figure 4. Comparative expression of genes for hematopoietic transcription factors. RNA was extracted from each cell line and gene expressions of hematopoietic transcription factors, such as Transcription factor PU.1 (PU.1), CCAAT/enhancer-binding protein alpha (CEBPα), Growth factor independent 1(GFI1), Neurogenic locus notch1 homolog protein 1 (NOTCH1), Transcription factor E2-alpha (E2A), GATA-binding factor 2 (GATA2), GATA1, and Friend of gata1 (FOG1) were investigated by real-time reverse transcription–polymerase chain reaction. All data were analyzed with the delta-delta Ct method.

Kasumi-1 cells that were established from acute and PU.1 expression. Our results also reveal hematopoietic myeloblastic leukemia (M2), reportedly express CD33, transcription factors CEBPα, PU.1 and GFI1 were expressed CD34, CD38, CD13, and CD15 (7), and the present study in Kasumi-1 cell in that order and their expressions are the confirms their expression. However, we also found lowest among the six cell lines we examined (Figure 4), expression of CD135, HLA-DR, CD45, CD14 and CD16 suggesting that Kasumi-1 cells could potentially be a useful with flow cytometric analysis, suggesting that we maintain tool for examining the effect of increasing the expression of this cell line as it is and our additional results offer robust these transcription factors. Using the surface markers that we evidence for defining these cells. As Kasumi-1 cells express examined here, we can expect to gain more evidence for Acute myeloid leukemia 1 (AML1)–Eight twenty one (ETO) differentiation. fusion gene product (13) and also have mutation in the c-KIT NB4 cells that were established from acute promyelocytic gene, they were utilized in research of regulation of the leukemia (M3) expressed CD38, CD33, CD13, CD15 and promoter region of the AML1–ETO fusion gene (14), of CD11b, but not CD11c, CD34, or CD14 (8). Except for inhibitors against c-KIT proteins (15) and of mechanisms of CD11c and CD14, our results shown in Figure 2 and 3B apoptosis (16) (17). Zapotocky et al. also focused on demonstrated the same phenotypic pattern as the previous hematopoietic differentiation and showed that valproic acid report. It is likely that this discrepancy was caused by triggers differentiation defined by c-Kit, CD33, CD34, difference in the antibody clones of CD11c and CD14 CD11b expression in Kasumi-1 cells. They also found the between our study and previous reports. In addition to the positive effect of valproic acid on up-regulation of CEBPα antibodies listed above, we also investigated the expression

4547 ANTICANCER RESEARCH 34: 4539-4550 (2014)

Table II. Summary of flow cytometric data.

Kasumi-1 Markers CD135– CD135+ CD135+ c-KIT- c-KIT+ c-KIT+ CD38+ CD38+ CD38– CD34+ CD34+ CD34– % 0.31±0.04 0.09±0.03 2.14±0.63 0.07±0.03 95.43±0.25 4.47±0.26 Markers HDR– HDR+ HDR+ HLADR– HLADR+ HLADR+ CD33+ CD33+ CD33– CD13+ CD13+ CD13– % 92.57±0.32 7.28±0.08 0.00±0.00 46.60±0.46 5.74±0.51 6.42±0.53 Among CD15– CD15+ CD15+ CD15– CD15+ CD15+ CD11b– CD11+ CD11+ CD13+ CD14+ CD14+ CD14– CD16+ CD16+ CD16– CD11c+ CD11c+ CD11c– CD116+ 0.69±0.18 0.13±0.07 1.99±0.43 0.50±0.08 0.05±0.00 1.93±0.50 0.02±0.01 0.12±0.04 14.86±0.61 NB4 Markers c-KIT– c-KIT+ c-KIT+ CD34– CD34+ CD34+ c-KIT– c-KIT+ c-KIT+ CD38+ CD38+ CD38– CD38+ CD38+ CD38– HLADR+ HLADR+ HLADR– % 42.13±0.71 14.37±3.50 10.23±1.28 28.83±2.57 2.61±0.24 7.43±0.37 16.93±3.61 2.66±0.65 6.71±0.88 Markers CD15– CD15+ CD15+ CD11b– CD11b+ CD11b+ CD11c– CD11c+ CD11c+ CD116+ CD116+ CD116– CD11c+ CD11c+ CD11c– CD14+ CD14+ CD14– % 13.20±2.86 1.96±0.14 3.15±0.33 5.39±0.71 21.17±2.87 56.63±2.38 9.26±0.33 21.17±1.35 48.03±1.12 MOLM-13 Markers HLADR– HLADR+ HLADR+ c-KIT– c-KIT+ c-KIT+ c-KIT– c-KIT+ c-KIT+ CD33+ CD33+ CD33– CD34+ CD34+ CD34– CD38+ CD38+ CD38– % 76.57±1.86 23.43±1.86 0.00±0.00 0.26±0.17 0.09±0.06 7.63±1.29 85.07±0.64 9.21±1.32 0.38±0.11 Markers HLADR– HLADR+ HLADR+ CD64– CD64+ CD64+ CD11b– CD11b+ CD11b+ CD13+ CD13+ CD13– CD13+ CD13+ CD13– CD13+ CD13+ CD13– % 2.16±0.14 0.85±0.10 21.73±0.42 1.37±0.12 0.01±0.01 0.02±0.00 1.56±0.13 0.04±0.01 0.23±0.01 Markers CD14– CD14+ CD14+ CD15– CD15+ CD15+ CD11c– CD11c+ CD11c+ CD116+ CD116+ CD116– CD13+ CD13+ CD13– CD14+ CD14+ CD14– % 0.39±0.07 0.10±0.02 0.12±0.03 1.42±0.18 0.18±0.05 7.79±1.63 2.22±0.63 0.01±0.01 0.01±0.01 MV4-11 Markers HLADR– HLADR+ HLADR+ c-KIT– c-KIT+ c-KIT+ CD38– CD38+ CD38+ CD33+ CD33+ CD33– CD34+ CD34+ CD34– CD34+ CD34+ CD34– % 14.80±0.78 85.03±0.67 0.00±0.00 0.14±0.06 0.03±0.01 6.02±0.54 0.14±0.06 0.05±0.01 0.63±0.03 Markers HLADR– HLADR+ HLADR+ CD14– CD14+ CD14+ CD15– CD15+ CD15+ CD13+ CD13+ CD13– CD116+ CD116+ CD116– CD64+ CD64+ CD64– % 2.76±0.18 2.62±0.36 30.30±2.67 0.36±0.05 0.02±0.01 0.13±0.06 0.35±0.10 30.70±3.70 57.93±1.16 Markers CD14– CD14+ CD14+ CD64– CD64+ CD64+ CD13– CD13+ CD13+ CD116+ CD116+ CD116– CD36+ CD36+ CD36– CD10+ CD10+ CD10– % 0.50±0.16 0.01±0.00 0.17±0.00 0.00±0.00 0.15±0.03 99.43±0.15 0.02±0.01 0.04±0.01 46.13±1.68 K562 Markers HDR– HDR+ HDR+ c-KIT– c-KIT+ c-KIT+ c-KIT– c-KIT+ c-KIT+ CD13+ CD13+ CD13– CD13+ CD13+ CD13– CD33+ CD33+ CD33– % 9.77±6.01 0.01±0.00 0.03±0.02 11.93±6.48 2.99±1.44 9.43±1.34 52.97±1.00 8.77±0.45 2.56±0.15 Markers c-KIT– c-KIT+ c-KIT+ CD38– CD38+ CD38+ c-KIT– c-KIT+ c-KIT+ CD34+ CD34+ CD34– CD33+ CD33+ CD33– CD36+ CD36+ CD36– % 0.42±0.18 0.10±0.06 8.51±0.32 57.70±1.14 0.02±0.01 0.00±0.01 57.90±2.09 17.37±1.42 5.43±0.50 Among c-KIT– CD36+ cells Markers CD71+ CD71– CD44+ CD44+ CD44– GPA+ GPA+ GPA– GPA+ GPA+ % 1.71±0.17 0.00±0.00 49.03±1.48 1.40±0.15 0.31±0.07 Jurkat Markers CD33– CD33+ CD33+ c-KIT– c-KIT+ c-KIT+ CD34– CD34+ CD34+ CD34+ CD34+ CD34– CD34+ CD34+ CD34– CD38+ CD38+ CD38– % 3.53±0.54 0.03±0.01 0.13±0.02 3.13±0.74 0.84±0.15 0.78±0.29 11.09±1.11 1.20±0.94 3.37±0.44 Among CD38+ CD7+ (1.50±0.40%) Markers CD2– CD2+ CD2+ CD3– CD3+ CD3+ CD3– CD3+ CD3+ CD5+ CD5+ CD5– CD4+ CD4+ CD4– CD8+ CD8+ CD8– % 0.00±0.00 1.48±0.40 0.01±0.00 0.12±0.04 0.67±0.18 0.62±0.16 0.00±0.00 0.01±0.01 1.18±0.31 Among CD38– CD7+ cells (59.43±2.40%) Markers CD2– CD2+ CD2+ CD3– CD3+ CD3+ CD3– CD3+ CD3+ CD5+ CD5+ CD5– CD4+ CD4+ CD4– CD8+ CD8+ CD8– % 0.36±0.39 56.73±2.55 2.26±0.07 6.84±0.31 21.82±1.12 23.44±1.18 0.01±0.01 0.03±0.01 39.01±2.60

4548 Inoue et al: Multicolor Analysis of Cell Surface Markers of Human Leukemia Cell Lines of c-KIT, HLA-DR, CD45, CD116 and CD16. NB4 cells cells (27). GATA2 and FOG1 expression was also confirmed have been utilized to examine the effect of differentiation (Figure 4) although their expression levels were lower than inducers (8, 18) and examined their effect by morphological GATA1, suggesting that GATA2 and FOG1 expression can be observation and surface expression utilizing CD11b used to trace the differentiation of K562 cells so as to antibodies. Although NB4 cells have been also used for evaluate the differentiation status more accurately. apoptosis studies (19, 20), no mention about surface marker Jurkat cells were established from acute T-cell leukemia expression has been made. Combined with studies on surface and not defined in terms of surface marker-based markers, these studies will provide us with more information differentiation state when they were established (12). Our on which cell fractions are most effective in apoptosis. results revealed that Jurkat cells are differentiated into Additional surface markers tested here, such as c-KIT, HLA- mature functional T-cells with mature T-cell markers (Figures DR, CD45, CD116 and CD16, support the robust evidence 2 and 3F), and T-cell-specific genes (NOTCH1 and E2A) to define the cells. were highly expressed (Figure 4), which implies that this cell MOLM-13 cells that were established from acute monocytic line is fit for studies that focus on transcriptional regulation, leukemia (M5a) expressed CD33, CD34, CD15, but not CD13 rather than differentiation studies. or CD14 (9), and our expression patterns for CD34 and CD14 The information mentioned herein can be used for differed compared with those previously reported (Figures 2 establishing standards for choosing appropriate cell lines for and 3C). This discrepancy also seemed likely to be due to functional analyses, and for determination of efficacy after differences in antibody clones of CD34 and CD14 between our functional analysis and drug screening. study and previous reports. There was no previous clarification of surface phenotype for MV-4-11 cells. p53-Dependent Acknowledgements apoptosis and cell cycle was investigated in MOLM-13 cells (21) (22) but without differentiation-based analyses. MOLM- We thank Dr. Kasem Kulkeaw, Ms. Tan Keai Sinn and Ms. Chiyoko 13 and MV-4-11 cells are both CD135- internal tandem Nakamichi for discussion, and the Ministry of Education, Culture, duplication (ITD)-positive AMLs, were utilized in CD135- Sports, Science and Technology, and the Ministry of Health, Labour and Welfare, and Japan Society for the Promotion of Science (JSPS) dependent drug-screening studies (23, 24) in which outcomes bilateral program for grant support. Tomoko Inoue was supported were evaluated with cell proliferation and phosphorylation of by a research award for young scientists from the Japan Leukaemia internal signal, such as CD135 itself and Signal transducer and Research Fund, a research grant from BioLegend/Tomy Digital activator of transcription 5 (STAT5), but not with surface Biology and Ogata memorial Science Foundation Research Fund. phenotype and transcription factor expression. Our results We thank the English editor for modifying our draft manuscript. shown in Figure 3C and D suggest that MOLM-13 cells and MV-4-11 are CFU-GM to myeloblasts and myeloblasts, References respectively, which implies that these cell lines will also be fit for functional analysis to examine differentiation after the 1 Okuhashi Y, Itoh M, Nara N and Tohda S: NOTCH knockdown myeloblast state to mature monocytes. Regarding gene affects the proliferation and mTOR signaling of leukemia cells. Anticancer Res 33: 4293-4298, 2013. expression, as CEBPα expression is quite high in MV4-11 cells 2 Tomiyasu H, Watanabe M, Sugita K, Goto-Koshino Y, Fujino Y, (Figure 4), reduction of expression would seem to be difficult Ohno K, Sugano S and Tsujimoto H: Regulations of ABCB1 and for this cell line, suggesting it is not suitable for studies of loss- ABCG2 expression through MAPK pathways in acute lymphoblastic of-function of CEBPα and its downstream genes. leukemia cell lines. Anticancer Res 33: 5317-5323, 2013. K562 cells are chronic myelogenous leukemia-derived 3 Inoue T, Sugiyama D, Kurita R, Oikawa T, Kulkeaw K, Kawano erythroleukemia cells and were not defined on surface H, Miura Y, Okada M, Suehiro Y, Takahashi A, Marumoto T, marker-based differentiation state when established (11); Inoue H, Komatsu N and Tani K: APOA-1 is a novel marker of erythroid cell maturation from hematopoietic stem cells in mice later studies confirmed the surface phenotype of GPA (25). and humans. Stem Cell Rev 7: 43-52, 2011. When examining the differentiation stage of erythroid cells, 4 Kulkeaw K, Inoue T, Mizuochi C, Horio Y, Ishihama Y and hemoglobinized cells were stained with benzidine and Sugiyama D: Ectopic expression of Hmgn2 antagonizes mouse hemoglobinization was quantified, depending on the staining erythroid differentiation in vitro. Cell Biol Int 36: 195-202, 2012. intensity. Recently Hu et al. reported that the distinct stages 5 De Rosa SC, Brenchley JM and Roederer M: Beyond six colors: in erythroid differentiation were clearly separated based on a new era in flow cytometry. Nat Med 9: 112-117, 2003. dynamic changes in membrane proteins, such as CD36, 6 Baumgarth N and Roederer M: A practical approach to CD44, CD71 and GPA using human cord blood-derived multicolor flow cytometry for immunophenotyping. J Immunol Methods 243: 77-97, 2000. erythroid cells (26). Herein, we confirm the expression of 7 Asou H, Tashiro S, Hamamoto K, Otsuji A, Kita K and Kamada these proteins in K562 cells. Concerning gene expression, N: Establishment of a human acute myeloid leukemia cell line GATA1 expression and its target gene effect was checked to (Kasumi-1) with 8;21 chromosome translocation. Blood 77: investigate its effect on erythroid differentiation in K562 2031-2036, 1991.

4549 ANTICANCER RESEARCH 34: 4539-4550 (2014)

8 Lanotte M, Martin-Thouvenin V, Najman S, Balerini P, Valensi F 20 Mozdarani H and Asghari F: Potent reducing effects of vitamin and Berger R: NB4, a maturation inducible cell line with D3 on the frequency of apoptosis induced by arsenic trioxide in t(15;17) marker isolated from a human acute promyelocytic NB4 cell line. Arch of Iran Med 13: 26-33, 2010. leukemia (M3). Blood 77: 1080-1086, 1991. 21 Kojima K, Konopleva M, Samudio IJ, Ruvolo V and Andreeff 9 Matsuo Y, MacLeod RA, Uphoff CC, Drexler HG, Nishizaki C, M: Mitogen-activated protein kinase kinase inhibition enhances Katayama Y, Kimura G, Fujii N, Omoto E, Harada M and Orita nuclear proapoptotic function of p53 in acute myelogenous K: Two acute monocytic leukemia (AML-M5a) cell lines leukemia cells. Cancer Res 67: 3210-3219, 2007. (MOLM-13 and MOLM-14) with interclonal phenotypic 22 Kojima K, Konopleva M, Tsao T, Nakakuma H and Andreeff M: heterogeneity showing MLL–AF9 fusion resulting from an Concomitant inhibition of Mdm2-p53 interaction and Aurora occult chromosome insertion, ins(11;9)(q23;p22p23). Leukemia kinases activates the p53-dependent postmitotic checkpoints and 11: 1469-1477, 1997. synergistically induces p53-mediated mitochondrial apoptosis 10 Lange B, Valtieri M, Santoli D, Caracciolo D, Mavilio F, along with reduced endoreduplication in acute myelogenous Gemperlein I, Griffin C, Emanuel B, Finan J, Nowell P and leukemia. Blood 112: 2886-2895, 2008. Rovera G: Growth factor requirements of childhood acute 23 Pauwels D, Klaassen H, Lahortiga I, Kilonda A, Jacobs K, leukemia: establishment of GM-CSF-dependent cell lines. Blood Sweron B, Corbau R, Chaltin P, Marchand A and Cools J: 70: 192-199, 1987. Identification of novel FLT3 kinase inhibitors. Eur J Med Chem 11 Lozzio CB and Lozzio BB: Human chronic myelogenous 63: 713-721, 2013. leukemia cell-line with positive Philadelphia chromosome. 24 Pietschmann K, Bolck HA, Buchwald M, Spielberg S, Polzer H, Blood 45: 321-334, 1975. Spiekermann K, Bug G, Heinzel T, Bohmer FD and Kramer OH: 12 Schneider U, Schwenk HU and Bornkamm G: Characterization Breakdown of the FLT3-ITD/STAT5 axis and synergistic of EBV-genome negative "null" and "T" cell lines derived from apoptosis induction by the histone deacetylase inhibitor children with acute lymphoblastic leukemia and leukemic panobinostat and FLT3-specific inhibitors. Mol Can Ther 11: transformed non-Hodgkin lymphoma. Int J of Cancer 19: 621- 2373-2383, 2012. 626, 1977. 25 Ma J, Hu Y, Guo M, Huang Z, Li W and Wu Y: hERG potassium 13 Kwong YL, Chan V, Wong KF and Chan TK: Use of the channel blockage by scorpion toxin BmKKx2 enhances polymerase chain reaction in the detection of AML1/ETO fusion erythroid differentiation of human leukemia cells K562. PLoS transcript in t(8;21). Cancer 75: 821-825, 1995. One 8: e84903, 2013. 14 Markova EN, Kantidze OL and Razin SV: Transcription of the 26 Hu J, Liu J, Xue F, Halverson G, Reid M, Guo A, Chen L, Raza AML1/ETO chimera is guided by the P2 promoter of the AML1 A, Galili N, Jaffray J, Lane J, Chasis JA, Taylor N, Mohandas gene in the Kasumi-1 cell line. Gene 510: 142-146, 2012. N and An X: Isolation and functional characterization of human 15 Shao HM, Tang YH, Shen Q, Zhu HQ, Ji O, Zhang YC, Ji JM, erythroblasts at distinct stages: implications for understanding Jiang PJ, Si YJ and Li ZR: Apoptosis of Kasumi-1 cells induced of normal and disordered erythropoiesis in vivo. Blood 121: by puerariae radix flavones and its molecular mechanism. 3246-3253, 2013. Zhonghua xue ye xue za zhi = Zhonghua xueyexue zazhi 33: 43- 27 Fujiwara T, Saitoh H, Inoue A, Kobayashi M, Okitsu Y, 46, 2012 (in Chinese). Katsuoka Y, Fukuhara N, Onishi Y, Ishizawa K, Ichinohasama R 16 Klampfer L, Zhang J, Zelenetz AO, Uchida H and Nimer SD: and Harigae H: 3-Deazaneplanocin A (DZNep), an inhibitor of The AML1/ETO fusion protein activates transcription of BCL-2. S-adenosylmethionine-dependent methyltransferase, promotes Proc Natl Acad Sci USA 93: 14059-14064, 1996. erythroid differentiation. J Biol Chem 289: 8121-8134, 2014. 17 Zapotocky M, Mejstrikova E, Smetana K, Stary J, Trka J and Starkova J: Valproic acid triggers differentiation and apoptosis in AML1/ETO-positive leukemic cells specifically. Cancer Lett 319: 144-153, 2012. 18 Laurenzana A, Petruccelli LA, Pettersson F, Figueroa ME, Melnick A, Baldwin AS, Paoletti F and Miller WH Jr.: Inhibition of DNA methyltransferase activates tumor necrosis factor alpha- induced monocytic differentiation in acute myeloid leukemia cells. Cancer Res 69: 55-64, 2009. 19 Han B, Ren Y, Guan L, Wei W, Hua F, Yang Y, Yang T, Cao T, Dong H, Pan H and Xu C: Sodium selenite induces apoptosis in Received April 4, 2014 acute promyelocytic leukemia-derived NB4 cells through Revised June 9, 2014 mitochondria-dependent pathway. Oncol Res 17: 373-381, 2009. Accepted June 10, 2014

4550